3956
Research Article
Cell organization, growth, and neural and cardiac
development require aII-spectrin
Michael C. Stankewich*, Carol D. Cianci, Paul R. Stabach, Lan Ji, Anjali Nath and Jon S. Morrow*
Department of Pathology, Yale University School of Medicine, 310 Cedar St. BML 150, New Haven, CT 06520, USA
*Authors for correspondence ([email protected]; [email protected])
Journal of Cell Science
Accepted 15 July 2011
Journal of Cell Science 124, 3956–3966
ß 2011. Published by The Company of Biologists Ltd
doi: 10.1242/jcs.080374
Summary
Spectrin a2 (aII-spectrin) is a scaffolding protein encoded by the Spna2 gene and constitutively expressed in most tissues. Exon trapping
of Spna2 in C57BL/6 mice allowed targeted disruption of aII-spectrin. Heterozygous animals displayed no phenotype by 2 years of age.
Homozygous deletion of Spna2 was embryonic lethal at embryonic day 12.5 to 16.5 with retarded intrauterine growth, and craniofacial,
neural tube and cardiac anomalies. The loss of aII-spectrin did not alter the levels of aI- or bI-spectrin, or the transcriptional levels of
any b-spectrin or any ankyrin, but secondarily reduced by about 80% the steady state protein levels of bII- and bIII-spectrin. Residual
bII- and bIII-spectrin and ankyrins B and G were concentrated at the apical membrane of bronchial and renal epithelial cells, without
impacting cell morphology. Neuroepithelial cells in the developing brain were more concentrated and more proliferative in the
ventricular zone than normal; axon formation was also impaired. Embryonic fibroblasts cultured on fibronectin from E14.5 (Spna22/2)
animals displayed impaired growth and spreading, a spiky morphology, and sparse lamellipodia without cortical actin. These data
indicate that the spectrin–ankyrin scaffold is crucial in vertebrates for cell spreading, tissue patterning and organ development,
particularly in the developing brain and heart, but is not required for cell viability.
Key words: Ankyrin, Neuroepithelial cell, Myocardium, Axons, Ventricular zone, Actin cytoskeleton, Cell growth, Development, Exon trap
Introduction
Spectrin a2 (aII-spectrin) is a 285 kDa scaffolding protein
abundant in most eukaryotic cells. Three decades of study have
implicated the spectrin heterodimer formed between aII-spectrin
and any of five b-spectrins in a bewildering array of cellular
processes. These include a role in the formation and maintenance
of specialized plasma membrane domains defining apical–
basolateral and planar polarity in epithelial cells, muscle and
neurons (Bennett and Baines, 2001); in the structural support of the
plasma membrane and the maintenance of cell shape (Gallagher
and Jarolim, 2000; Kizhatil et al., 2007); as a scaffold upon
which calcium-mediated and tyrosine kinase–phosphatase signal
transduction pathways converge (Nicolas et al., 2002; Nedrelow
et al., 2003); as a tumor-suppressor protein involved in TGF-b–
SMAD regulation (Tang et al., 2003); as a cargo selection
mechanism in the secretory and endocytic pathways (De Matteis
and Morrow, 2000); as a regulator of macropinocytosis (Xu et al.,
2000); as a tether linking trafficking vesicles to microtubule
motors (Holleran et al., 2001; Muresan et al., 2001); as a nuclear
scaffold organizer (McMahon et al., 1999; Tse et al., 2001); and
most recently, as a potential mechano-sensing ligand-binding
switch (Stabach et al., 2009). Deletion of aII-spectrin in
Drosophila melanogaster and Caenorhabditis elegans leads to
late embryonic–early larval stage lethality (Moorthy et al., 2000;
Dubreuil, 2006; Hammarlund et al., 2007), and recent knockdown
studies of aII-spectrin in cultured cells have demonstrated growth
and adhesion defects (Metral et al., 2009). However, the role of
aII-spectrin in vertebrate development remains unexplored.
We have achieved targeted disruption of aII-spectrin in
C57/B6 mice by the insertion of a foreign exon encoding
b-galactosidase (b-gal) into the murine Spna2 gene. The resulting
gene product is a short-lived and non-functional fusion protein
that includes the N-terminal half of aII-spectrin fused to b-gal.
Heterozygous animals (Spna2+/2) display no phenotype by
2 years of age. Homozygous animals die in utero. Embryos
lacking wild-type aII-spectrin (Spna22/2) are smaller than
normal, survive to embryonic day 12.5 (E12.5), but then die
due to malformations of the neural tube and cardiac systems.
Thus, aII spectrin is required in vivo for late embryonic
development. The cell biological consequences of its loss
include instability of its cognate partners (bII- and bIIIspectrin); impaired membrane biogenesis and sorting of not
only the cortical actin skeleton, but also of ankyrins B and G; and
modification of pathways regulating cell spreading and growth.
Conversely, spectrin loss in vivo does not affect epithelial cell
shape, stability or nuclear morphology, nor is it required for
cell viability. These findings have several implications for the
potential role of aII-spectrin in human disease.
Results
Loss of aII-spectrin leads to a concomitant loss of two
b-spectrins and selected ankyrins
PCR analysis of murine tissues derived from ES RRQ171
identified the locus of the b-geo exon trap (encoding a b-gal–
neomycin resistance fusion protein) within the intron between
exons 24 and 25 of the spectrin Spna2 gene (Fig. 1A). The exontrapped gene generates a spectrin b-gal fusion message that
truncates the aII-spectrin gene product at codon 1153,
corresponding to a polypeptide terminating within spectrin
repeat ten, lacking the C-terminal site responsible for
Deletion of aII-spectrin
3957
Journal of Cell Science
Fig. 1. Targeted disruption of aII-spectrin. (A) An ES
cell line was established using gene trap vector RRQ171
(b-geo) from Bay Genomics. Analysis by 59 RACE
identified insertion of the vector between exons 24 and 25 of
the murine Spna2 gene. This created a fusion transcript with
a spectrin message truncated by the addition of b-geo. A
cartoon of this fusion transcript, and the anticipated fusion
protein, is depicted. (B) E11.5 embryos derived from a
RRQ171 heterozygous breeding pairs were genotyped by
quantitative RT-PCR for b-geo. (C) RT-PCR analysis with
intron-bridging primer pairs directed to upstream exons (6/7)
or downstream exons (54/55) confirmed the absence of
mRNA encoding full-length aII-spectrin in homozygotes.
(D) Western blot analysis showed that monoclonal
antibodies to aII-spectrin (aII-C) that react with peptide
sequences downstream of the exon trap were negative in
homozygotes. Pan-reactive anti-spectrin antibodies (aII-pan)
detect the fusion protein. Antibodies to b-gal confirm the
presence of the fusion protein in both the homoand heterozygotes.
heterodimer formation with b-spectrin (Li et al., 2008). In tissues
or cells homozygous for this insertion (Fig. 1B), mRNA
encoding aII-spectrin was undetectable when probed by
realtime (RT)-PCR for sequences downstream of b-geo, but not
when primers upstream of exon 24 were utilized (Fig. 1C).
Correspondingly, western blot analysis showed that Spna22/2
embryos did not react with aII-spectrin antibodies directed to
epitopes downstream of the exon trap, whereas antibodies
directed to epitopes upstream of the exon trap reacted with a
band at about 260,000 relative molecular mass (Fig. 1D). This
band also reacted with antibodies against b-gal, confirming its
identity as a fusion product between aII-spectrin exons 1–24 and
b-gal. Analysis of whole embryos for the expression of either the
wild-type gene or the exon-trapped gene revealed no significant
differences, with both being widely expressed and most highly
expressed in the nervous system (supplementary material
Fig. S1). By densitometry analysis, the ratio of the spectrin–bgal fusion protein in Spna22/2 mice to aII-spectrin in wild-type
embryos (normalized to actin) was 0.21±0.03 (±1 s.d. arbitrary
units). Analysis of mRNA levels indicated that the mRNA for
wild-type aII-spectrin and the mRNA for the fusion product were
expressed at equal levels (Fig. 2B). Thus, the spectrin–b-gal
fusion protein must be unstable relative to the wild-type protein;
and the transcription level of mRNA encoding aII-spectrin is not
responsive to the loss of the mature protein.
Also of interest was the fate of other proteins that are typically
tightly associated with aII-spectrin, specifically spectrins bII and
bIII, and ankyrins R, G, and B. Whole embryo levels of bII- and
bIII-spectrin were reduced to below 20% of normal in the
absence of aII-spectrin (Fig. 2A), with no change in their level of
mRNA expression (Fig. 2B). This finding is consistent with
earlier studies in vertebrates demonstrating that b-spectrin is
degraded if not assembled with a-spectrin (Woods and Lazarides,
1985; Hanspal and Palek, 1987), and also indicates an absence of
direct feedback control on the levels of b-spectrin expression.
The level of spectrins aI and bI, forms predominantly (albeit not
exclusively) expressed in erythrocytes, were unchanged in the
Spna22/2 mice, at either the protein or mRNA level. Neither of
the two remaining spectrins (bIV and bV) were expressed by
E14.5 at sufficient levels to enable their reliable detection, so the
impact of aII-spectrin loss on the disposition of these proteins
was not evaluated.
Ankyrin expression and accumulation was less severely
perturbed compared with the spectrins, but the reductions were
Journal of Cell Science 124 (23)
Journal of Cell Science
3958
Fig. 2. Loss of aII-spectrin destabilizes bII- and bIII-spectrin.
(A) Western blot analysis of whole embryos comparing the relative
steady state protein levels of several spectrins and ankyrins. Each lane
represents results with a separate embryo. The loss of aII-spectrin
reduced the steady state levels of bII- and bIII-spectrin to below 20% of
normal; aI- and bI-spectrin were unchanged. The abundance of bIV- and
bV-spectrin was too low in embryos younger than E14.5 to be reliably
evaluated. Ankyrins B440 and G190 were both significantly diminished,
as were ankyrins B220 and B150, albeit not to the same degree as the
b-spectrins. (B,C) Quantitative RT-PCR analysis revealed that despite
the change in protein levels, there were no consistent changes in the
mRNA levels of any spectrin or ankyrin (except for the disrupted Spna2
gene, as measured with primers targeted to the 39 end, aII-39). Each
analysis was performed in triplicate on two or three separate animals.
Error bars show ±1 s.d. **P,0.05, ***P,0.005.
still significant (Fig. 2A,C). The protein levels of AnkG190
(P,0.005) and AnkB440/220 (P,0.05) were slightly reduced and
no significant changes in levels of mRNA encoding ankyrin were
detected. The overall levels of VASP (Bournier et al., 2006), Abi
(Hssh3bp1) (Ziemnicka-Kotula et al., 1998), Kap3 (Takeda et al.,
2000), and 14-3-3 (Ramser et al., 2010), proteins that directly bind
aIIbII-spectrin, were unchanged (supplementary material Fig. S2).
aII-spectrin deficiency is embryonic lethal, with cardiac,
craniofacial and neural tube malformations
The loss of aII-spectrin is embryonic lethal. Of 18 litters generating
127 C57BL/6 mice in matings between heterozygous Spna2+/2 pairs,
68 heterozygous and 27 homozygous wild-type mice were produced.
No homozygous Spna22/2 mice were born. Heterozygous
(Spna2+/2) mice were morphologically indistinguishable from
wild-type animals. Thus, one normal Spna2 allele was sufficient
and generated normal levels of aII-spectrin (Fig. 1).
Examination of the homozygous embryos at different
gestational ages indicated that most (but not all) such embryos
were still viable at E12.5, but that all had died by E16.5; thus the
loss of aII-spectrin causes embryonic death between E12.5 and
E16.5. Homozygous embryos exhibited intrauterine growth
retardation; the average length of the Spna22/2 embryo at
E11.5 was approximately 5.5 mm, and 7.0 mm for wild-type
littermates. Whole-mount examination of heterozygous embryos
revealed widespread expression of aII-spectrin throughout most
tissues, but most concentrated in the heart and nervous system.
The expression pattern of the mutant aII-spectrin–b-gal fusion
allele was similar to that of the wild-type protein (supplementary
material Fig. S1).
Accompanying the lethal phenotype of the Spna22/2 animals
were several gross morphologic abnormalities (Fig. 3). Although
not every homozygous embryo showed every abnormality,
typical features associated with the Spna22/2 embryos included
craniofacial abnormalities with incomplete neural tube closure;
abnormal dilatation of the primary brain vesicles, constriction of
the mesencephalic vesicle, along with lateral ventricle and
forebrain anomalies (Fig. 3B). Outside the nervous system, aIIspectrin-deficient animals exhibited abnormal cardiac shape,
cardiac dilatation, and thinning of the compact myocardium
Journal of Cell Science
Deletion of aII-spectrin
3959
Fig. 3. Spna22/2 embryos die at E12.5–16.5 with
multiple defects. (A,B) Whole mounts depicting the
gross morphological defects in head development,
with frequent neural tube closure defects (E11.5
embryos shown). In cases where the neural tube
closed, there remained distinct alterations in head
and back curvature with craniofacial abnormalities.
(C) Serial sections through the fetal heart. The
Spna22/2 hearts have an irregular shaped ventricle,
and a thinned compact myocardium. Scale bar:
100 mm. (D) Quantitative comparison of cardiac wall
thickness. On average, the myocardium of the
Spna22/2 hearts was 70.9% of the thickness of normal
hearts of the same gestational age. This difference is
highly significant (n5101; ***P54.2610213). Error
bars indicate s.d.
(Fig. 3C). The cardiac changes probably initiated the embryonic
lethality, and were reminiscent of cardiac phenotypes observed
following the disruption of TGF-b–Smad signaling (Qi et al.,
2007) or following the loss of proteins that regulate actin
dynamics such as Ena/VASP (Eigenthaler et al., 2003).
Interestingly, both TGF-b–Smad signaling (Tang et al., 2003)
and Ena/VASP (Bournier et al., 2006) interact with spectrin,
although the levels of neither Ena/VASP (supplementary material
Fig. S2) nor the Smad proteins (supplementary material Fig. S3)
were altered in the Spna22/2 embryos (see Discussion). Two
unique cardiac embryonic isoforms of aII-spectrin (aIIS9 and
aIIS10 ) that arise by alternative splicing have also been recently
reported (Zhang et al., 2010). In cell culture, these isoforms
promote myocyte growth, suggesting that their loss in the
Spna22/2 animals is consistent with the cardiac phenotype
observed. Finally, the cardiac changes appear not to be secondary
to failed hematopoiesis because the embryonic blood islands
appeared normal and the embryos were pink and well
oxygenated.
At the histological level, Spna22/2 animals displayed changes
in the density of neuroepithelial cells along the ventricular zone
(VZ) of the developing brain (Fig. 4A,B). This is a region of
active neuronal and glial differentiation; cells from this zone
normally proliferate and migrate apical to populate overlying
layers (Gotz and Huttner, 2005). At about E13.5, earlier work has
demonstrated that bII-spectrin and ankyrin B mark axonal sprouts
in developing neuroblasts, whereas ankyrin G is a reliable marker
of nascent axonal initial segments (Tang et al., 2002). These
distributions can be appreciated in the Spna2+/+ animals, as
shown in Fig. 4B. In the spectrin-deficient animals, these
markers became restricted to an area near the soma, and the
developing axonal sprouts and initial axon segments were
diminished in abundance, a finding also supported by the
reduced tau staining revealed in Fig. 4C. These animals also
displayed enhanced nuclear density in the subventricular zone
(SVZ), with more cells in the Spna22/2 animals staining positive
for the proliferation marker Ki67 (34±11% vs 46±8%, P,0.006).
There was no change in the level of apoptosis (as measured by
TUNEL staining; Fig. 4C). Beyond greater cellular proliferation,
and given the impaired cellular spreading and movement of aIIspectrin-deficient cells described below and by others (Metral
et al., 2009), the enhanced cellular density apparent in the VZ and
SVZ of the Spna22/2 animals could also arise from a failure
of interkinetic movement of nuclei in the neuroepithelial cells
of this zone, a feature related to cellular proliferation and
responsible for the pseudostratified appearance of nuclei in the
VZ (Gotz and Huttner, 2005). Finally, using tau as a marker of
axonal processes, it is apparent that aII-spectrin deficiency
impaired axon formation and/or axon guidance in the VZ
(Fig. 4C). Tau-positive axons were shorter, less ordered, and
often truncated at or near the soma in these animals. In this same
area in the spectrin-deficient animals there was enhanced
vimentin staining, a marker of glia, (supplementary material
Fig. S4), suggesting that, as with the loss of bII-spectrin, aIIspectrin deficiency altered the program of neuroepithelial cell
differentiation (Golestaneh et al., 2006). Earlier work has
Journal of Cell Science 124 (23)
Journal of Cell Science
3960
Fig. 4. Histology of Spna22/2 embryos. (A) The VZ and SVZ of the developing brain display significant histological changes. Wild-type embryos at this stage
of development display a thick pseudostratified VZ with active upward migration and population of overlying layers from the SVZ. In the Spna22/2 animals, the
VZ was more compact, with fewer cells filling the more superficial layers (boxed area, shown enlarged in B). Other organs revealed no significant histological
changes; kidney and lung are shown. In these organs, tubules and bronchioles were developing normally. (B) Enlargement of area outlined in A. In wild-type
animals, aII- and bII-spectrin are rich in axon bundles (boxed) and are also present at synaptic terminals and initial axon segments (arrows). Ankyrin B is
associated with unmyelinated axonal processes in the embryonic mouse (Chan et al., 1993), whereas ankyrin G is found along axons and in dendrites and
concentrates in the initial axon segments (Kordeli et al., 1995). The loss of aII-spectrin disrupts the localization of these proteins, especially along the axon
bundles, which no longer are marked by either spectrin or ankyrin. (C) Top two rows: Staining of sections with anti-tau demonstrates developing axon bundles in
the VZ. Magnified images (36) of the boxed areas are shown on the right with inverse contrast. Note the foreshortened, sparse and more disordered axon
segments and increased concentration of tau near the soma when aII-spectrin is disrupted. Bottom two rows: Immunostaining for Ki67, a proliferation marker,
reveals increased proliferation with spectrin loss. Averaged over multiple fields, and scoring any detected Ki67 staining as positive, 34±11% of cells in the wildtype animals were positive as compared with 46±8% of the cells in the spectrin-deleted animals. This represents a highly significant (P,0.006) 35% increase in
proliferative activity. TUNEL staining (right) reveals no change in the level of apoptosis. (D) Higher power views of developing renal tubules and lung
bronchioles indicate preservation of epithelial morphology. Comparison of epithelial cell height reveals no changes, even without laterally associated spectrin or
ankyrin (see Fig. 5).
demonstrated that the micro-injection of anti-spectrin antibodies
can lead to a condensation of the vimentin network (Mangeat and
Burridge, 1984). It will thus be of interest in future studies to
carefully examine the vimentin network in cultured spectrindeficient cells because alterations in vimentin might contribute to
some of the cell growth and phenotypic changes observed.
In contrast to the histological changes apparent in the brain, no
histological anomalies were detected in E13.5 embryos in several
other tissues. Sections of the developing kidney and lung are
shown in Fig. 4A,D. Although spectrin and ankyrin presumably
play a role in providing structural support to the plasma
membrane, and their absence in erythrocytes (Gallagher and
Jarolim, 2000) and isolated bronchial cells (Kizhatil et al., 2007)
leads to profound cell-shape change, the loss of aII-spectrin (with
concomitant loss of bII- and bIII-spectrin) resulted in no
significant change in epithelial cell layer thickness or epithelial
cell shape (Fig. 4D). Quantitative evaluation of cells with welldefined borders found no significant differences in cell height
Deletion of aII-spectrin
[6.46±1.1 versus 6.5±1.0 (±2 s.d. arbitrary units)] between
Spna22/2 and wild-type embryos.
Spectrin loss causes a redistribution of bII- and
bIII-spectrin and ankyrin
unlike normal aII-spectrin (that surrounds the entire membrane).
Interestingly, both distributions were apparent in heterozygous
animals, suggesting that the fusion protein distributes
independently of and has no effect on normal spectrin. This
was also apparent when cell extracts were sedimented (Fig. 5A).
In wild-type or heterozygous animals, aII-spectrin sedimented as
a large multiprotein complex near 13S, as did bII-spectrin.
In animals without aII-spectrin, the residual bII-spectrin
sedimented near 8S, indicating that it is no longer retained in a
high molecular weight multiprotein complex.
The bII- and bIII-spectrin that remained in the renal tubules
and lung bronchioles was largely redistributed to an apical
submembranous compartment, coincident with the aII-spectrin–
b-gal fusion protein (Fig. 5B). Ankyrin G displays the same
redistribution; ankyrins B and R are only very weakly expressed
in these cells at this stage of development and did not stain
reliably. A consideration in these results is whether the absence
of intact aII-spectrin per se is the driving force for protein
redistribution, or whether it is an epiphenomenon due to a strong
sorting determinant on the aII-spectrin–b-gal fusion protein
present in these cells. Although we cannot formally exclude this
possibility, we do not favor it, given the normal distribution of
bII-spectrin and ankyrins in the heterozygote (data not shown).
Journal of Cell Science
The loss of aII-spectrin perturbs the levels and distribution of its
associated proteins. In the VZ at E13.5, aII- and bII-spectrin
normally are found along the developing axon bundles (Fig. 4B).
Ankyrins B and G display a similar distribution, while ankyrin G
also concentrates in the initial axon segments. In the absence
of aII-spectrin, bII-spectrin and ankyrin B are lost from the
unmyelinated axonal processes and ankyrin G is no longer
concentrated at initial segments.
Cells outside the nervous system also show a redistribution
of the proteins normally associated with aII-spectrin, although
apical–basolateral polarity is retained. Results with kidney
tubules and lung bronchioles are shown (Fig. 5). As expected,
in the Spna22/2 animals, there was no immunostaining of tissues
with antibodies directed to the C-terminal portions of aII-spectrin
(Fig. 5A). When antibodies were used that recognize the spectrin
portions of the spectrin–b-gal fusion, or with b-gal antibodies,
the fusion product was found concentrated along the apical
membrane and submembrane in epithelial cells, a distribution
3961
Fig. 5. Loss of aII-spectrin leads to a redistribution of both bII- and bIII-spectrin and ankyrin. (A) In developing renal tubules, aII-spectrin is distributed
uniformly over both the basolateral and apical surfaces in wild-type (Spna2+/+) embryos, as detected by both the C-terminal anti-aII-spectrin antibody (aII-C) as
well as the pan-reactive spectrin antibody (aII-pan). In the Spna22/2 tubules, the residual spectrin–b-gal fusion protein is totally lost from the basolateral
membrane, and concentrated in coarse apical and sub-apical pools as detected by both aII-pan and by an antibody to b-gal. In heterozygotes (Spna2+/2), the
presence of the fusion protein had no detectable effect on the distribution of wild-type aII-spectrin, nor was the distribution of the fusion protein changed by the
presence of aII-spectrin. Below: Sucrose density gradient analysis of proteins extracted from E13.5 embryos. In the homozygote (Spna22/2), bII-spectrin does not
sediment with the high molecular weight protein complex at 13S. (B) Immunofluorescent micrographs of bronchioles and renal tubules. Without intact aIIspectrin, bII- and bIII-spectrin and ankyrin G are largely absent from the lateral membrane of both renal and bronchiole epithelial cells. Residual bII- and bIIIspectrin and ankyrin G concentrate instead at the apical membrane in a pattern similar to that of the aII-spectrin–b-gal fusion protein.
3962
Journal of Cell Science 124 (23)
Journal of Cell Science
Fig. 6. MEFs lacking intact aII-spectrin
display impaired growth and spreading, and
altered morphology. (A,B) MEFs from E14.5
embryos were plated onto fibronectin, grown for
up to 3 hours, and stained as indicated. The
images shown here were from cells examined at
30 minutes after replating. (A) MEF from wildtype cells, stained for F-actin with rhodamine
phalloidin. Note the peripheral cortical and
perinuclear actin, cell spreading, and abundant
ruffling edges and lamellipodia. (B) Same wildtype cells as in A, stained for aII-spectrin. Note
the coincidence of spectrin with the ruffling
edge (arrows). (C,D) MEFs from Spna2+/+
(C) or Spna22/2 (D) embryos under same
conditions as in A, stained for actin. Note the
lack of spreading and lamellipodia, and
abundant filopodial projections and stress
filaments in the spectrin-deficient cells.
(E–H) MEFs from Spna22/2 embryos under
same conditions as A, but transfected with wildtype GFP-labeled aII-spectrin. Double
immunofluorescent images are shown, stained
for actin or GFP. Arrows mark the accumulation
of GFP–spectrin at the ruffling edge and
lamellipodia. Note the recovery of cortical and
perinuclear actin, ruffling, and spreading in the
transfected cells expressing GFP–aII-spectrin.
(I) Scatter diagram of individual cell areas of
MEFs from Spna2+/+ (WT) or Spna22/2 (KO)
embryos, or from Spna22/2 cells transfected
with GFP–aII-spectrin (KO+aII). The KO cells
are significantly less spread, with an average
area just 32% of wild-type cells
(***P#1.6610216); there is no difference in
area between the wild-type cells and the
Spna22/2 cells transfected with GFP–aIIspectrin. (J) Comparison of cell growth in
culture, as measured by cell count, of wild-type
and two MEF lines (KO1, KO2) independently
derived from separate Spna22/2 embryos.
We also note that in Drosophila, the genetic loss of a-spectrin
leads to a redistribution and mislocalization of b-spectrin (Garbe
et al., 2007).
Growth and cell spreading is impaired in Spna22/2 MEFs
To determine the effect of spectrin loss on cellular actin and cell
shape, the morphology of mouse embryonic fibroblasts (MEF)
cultured from the Spna22/2 animals at E14.5 was compared with
wild-type cells harvested from matched littermates. MEFs were
plated on fibronectin and after 60 minutes, wild-type MEFs
displayed a spread and flattened cell shape, with cortical and
perinuclear actin bundles and frequent lamellipodia that were rich
in both actin and aII-spectrin (Fig. 6A–C). The aII-spectrin null
MEFs spread more slowly, and after 0.5–3 hours in culture,
displayed many filopodia and stress fibers, but few lamellipodia,
giving them a rounded and spiky appearance (Fig. 6D). These
cells also lost their cortical actin ring, as well as the actin
pools surrounding the nucleus. This phenotype was rescued
by transfection with wild-type EYFP-tagged aII-spectrin, with
restoration of a spread morphology, ruffling edge, and cortical
actin (Fig. 6E–H). The expressed EYFP–aII-spectrin was
apparent in the rescued cells at the ruffling edge and in
lamellipodia, in a pattern similar to spectrin in wild-type cells
(arrows, Fig. 6F,H). The spread area of the aII-spectrin-deficient
MEFs was on average only 32.2±20% of the area of the wild-type
cells (Fig. 6, scattergram). This result was highly significant
(P51.6610216 by ANOVA). This effect on cell spreading and
size was completely rescued by transfection with EYFP–aIIspectrin.
Surprisingly given the enhanced Ki67 labeling observed in the
brain of the mouse embryos, MEFs lacking aII-spectrin grew more
slowly than did wild-type cells (Fig. 6, bottom). This reduction in
cell proliferation was not due to increased apoptosis or oncosis
because the fraction of unfixed cells permeant to propidium iodide
or the level of annexin V staining was unchanged between wildtype and spectrin deficient cells (data not shown). Whereas the
basis for this difference between in vivo proliferation and that
observed in culture is unknown, these findings do point to a
potentially significant role for spectrin in modulating the signals
that control cell growth and differentiation.
Journal of Cell Science
Deletion of aII-spectrin
Discussion
Exon trapping has been used to generate a strain of mice
devoid of functional aII-spectrin. These mice display uniform
embryonic lethality at E12.5 to E16.5, with gross craniofacial,
neurodevelopmental and cardiac defects. Detailed analysis
reveals: (1) no compensation by aI-spectrin for the loss of aIIspectrin; (2) a reduction of bII- and bIII-spectrin to about 20% of
normal levels due to accelerated degradation, with no change in
their mRNA levels; (3) no apparent hematological anomalies that
might contribute to the cardiovascular defects; (4) preserved
epithelial cell morphology in renal tubules and bronchial
epithelium; (5) mis-sorting of residual bII- and bIII-spectrin
and ankyrin G in epithelial cells of the kidney and lung, away
from the basolateral membrane to an apical compartment; (6)
abnormal cellularization of the ventricular zone of the developing
brain, with impaired axon formation; and (7) impaired spreading
and lamellipodia formation, and reduced proliferation, in MEF
cells cultured from aII-spectrin-deficient animals.
The phenotype observed in these animals is very similar,
although not an exact phenocopy, to that observed in mice
deficient in bII-spectrin (termed ELF in those studies) (Mishra
et al., 1999; Tang et al., 2003; Golestaneh et al., 2006). The loss
of bII-spectrin appears to disrupt growth and differentiation
control via the TGF-b and Smad signaling pathways, and in this
context bII-spectrin acts as a Smad3/4 adapter tethering these
proteins to TFG-b receptors at the membrane (Tang et al., 2003).
Given that the loss of aII-spectrin leads to a secondary loss of
most bII- (and bIII)-spectrin, it is tempting to attribute the
phenotype of the aII-spectrin null animals to a disruption of bIIspectrin. However, several features of the Spna22/2 animals
indicate that their phenotype is subtly different.
The loss of TGF-b signaling in bII-spectrin null animals leads
to not only cardiac and neurological defects, but also to
intrahepatic biliary tree anomalies and the late onset of
gastrointestinal and hepatic tumors in heterozygotes (Tang
et al., 2005). The aII-spectrin mutant animals display no
hepatic anomalies, at least during early embryonic
development, and after 2 years the heterozygotes show no
propensity to tumor formation. Similarly, both the aII- and bIIspectrin null animals display disrupted neuroepithelial cell
compaction in the ventricular zone (Golestaneh et al., 2006)
and both display an exaggerated proliferation of neuroepithelial
precursors, suggesting that aII-spectrin-deficient animals might
also harbor a defect in TGF-b-mediated cell differentiation.
Although this result might be expected based on the reduced
levels of bII-spectrin that follow the loss of aII-spectrin, we have
so far been unable to demonstrate any changes in Smad2/3 levels
or in their levels of phosphorylation in aII-spectrin-deficient
animals or cells (supplementary material Fig. S3). Thus, possibly
only low levels of residual bII-spectrin are sufficient to maintain
TGF-b–Smad signaling pathways. This result would imply that
bII-spectrin can function autonomously from aII-spectrin. There
is some evidence for autonomous b-spectrin function, beginning
with early observations of homopolymeric b-spectrin associated
with the neuromuscular junction (Bloch and Morrow, 1989;
Pumplin, 1995), and further supported by studies in Drosophila
that have revealed persistent b-spectrin function in axonal
pathfinding that is partially independent of a-spectrin (Garbe
et al., 2007; Hulsmeier et al., 2007).
Alternatively, other changes more specific for aII-spectrin
might be the genesis of the observed changes in the heart and
3963
brain. In many respects, the aII-spectrin null animals appear
similar to those with deletion of n-cofilin (Bellenchi et al., 2007),
which causes impaired radial migration and early compaction
of the VZ due to an early exit of progenitor cells from
the proliferating pool. Impaired spreading (and impaired
lamellipodia formation) is a major consequence of aII-spectrin
deficiency in the isolated cell culture studies reported here, as
well as in earlier aII-spectrin knockdown studies (Metral et al.,
2009). aII-spectrin was also found to interact via its SH3 domain
with EVL/VASP, an actin regulator (Bournier et al., 2006). The
importance of actin dynamics as a regulator of radial migration
from the VZ is also supported by the observation that mutations
in filamin A (another actin-binding protein) profoundly affect
neuroepithelial polarity and migration from the VZ (Sato and
Nagano, 2005). Thus, we speculate that the subtle phenotypic
differences observed between the deletion of aII-spectrin and bIIspectrin reflects in part, a more widespread disruption of actin
dynamics, with perhaps less contributions from impaired TGF-b–
Smad signaling.
The ab-spectrin heterotetramer also has other putative
functions beyond actin binding that might contribute to the
phenotype of its deficiency. The SH3 domain of aII-spectrin is
involved in macropinocytosis (Xu et al., 2000) and cell signaling;
it binds both non-receptor tyrosine kinases of the c-src family
(Nedrelow et al., 2003) and tyrosine phosphatases (Lecomte
et al., 2001). The tyrosine phosphorylation of spectrin blocks its
cleavage by calpain (Nicolas et al., 2002; Nedrelow et al., 2003),
an effect also modulated by Ca2+ and calmodulin (Harris and
Morrow, 1990). Spectrin participates in the activation of the Rho
GTPase Rac, and overexpression of the aII-spectrin SH3 domain
inhibits Rac1 activation, actin filament formation, and cell
spreading in cultured cells (Bialkowska et al., 2005).
Interestingly, actin and Rac1 have also been implicated in the
regulation of neuronal polarity and axon formation via the
WAVE complex (Tahirovic et al., 2010), and the loss of Rac1 in
neural crest cells leads to severe craniofacial and cardiovascular
malformations (Thomas et al., 2010). We thus anticipate that a
major effector pathway disrupted in the Spna22/2 animals will be
Rac-mediated actin regulation.
A third aspect of spectrin’s function that might contribute to
the observed phenotype is its participation as a membrane
stabilizer. Studies in C. elegans (Hammarlund et al., 2007) and
Drosophila (Garbe et al., 2007; Hulsmeier et al., 2007)
demonstrate aberrant neurite development and axon guidance,
and (in the case of C. elegans) frequent axonal breaks with
chaotic repair in spectrin-deficient animals (Hammarlund et al.,
2007). These observations build on the classical view of the
spectrin skeleton as a stabilizing and organizing infrastructure.
However, the evidence for this process in vertebrate nonerythroid cells is scant. Only in mammalian erythrocytes is
the spectrin–ankyrin membrane skeleton continuous and
homogeneous. In other cells, including neurons, wide variations
exist within the cell in terms of the membrane organization
of spectrin or ankyrin, both in terms of its uniformity and
composition, and in whether spectrin and/or ankyrin are even
associated with the plasma membrane or localized on internal
compartments (De Matteis and Morrow, 2000). When spectrin is
concentrated at the membrane, it is most often at sites of
specialized receptor function such as pre- and post-synaptic
densities, rather than at sites that one would anticipate are
subjected to high mechanical stress. However, recent data
Journal of Cell Science
3964
Journal of Cell Science 124 (23)
suggesting that spectrin might function as a mechanochemical
signal transducer (Stabach et al., 2009) raises a speculative but
interesting possibility that could reconcile the apparent fragility
of spectrin-deficient axons (as observed in C. elegans) with the
emerging role of spectrin as a major receptor-sorting and
organizing scaffold. If stretch or bending of axons is a tropic
stimulus for neuronal extension and maintenance (Van Essen,
1997), then the absence of such stimulus might activate pathways
to prune unwanted axons and dendrites (Luo and O’Leary, 2005).
If indeed a spectrin lattice is part of a mechanosensing pathway
by which tension, stretch or bending is transduced to trophic
signals, the absence of spectrin might leave neurons poised for
degradation and pruning not by a mechanical instability, but
rather because of an absence of maintenance signals.
Finally, it is worthwhile to consider the contributions of
another major function of the spectrin–ankyrin skeleton to the
pathology of the Spna22/2 animals. Based on studies in cultured
cells, gene deletion studies, and linkage analysis of human
diseases, it is well established that a key function of the spectrin–
ankyrin skeleton is to facilitate the movement of selected
membrane proteins through the secretory and endocytic
pathways to points of physiologic action (reviewed in De
Matteis and Morrow, 2000; Bennett and Healy, 2008). Thus,
patients with mutations in ankyrin B mis-sort voltage-gated
sodium channels in heart muscle suffer from long QT syndrome
type 4 and sudden cardiac death. Many other proteins also depend
on ankyrin for efficient sorting, including the IP3 receptor (Tuvia
et al., 1999) and aI-Na,K-ATPase (Stabach et al., 2008). In
lymphocytes, the trafficking of receptors CD3 and CD45 are both
spectrin- and ankyrin-dependent (Pradhan and Morrow, 2002).
Mutations in bIII-spectrin cause spinocerebellar ataxia type 5 in
humans (Ikeda et al., 2006) (and ataxia and seizures in mice)
(Perkins et al., 2010; Stankewich et al., 2010) due to impaired
intracellular transport of EAAT4 and at least five other proteins
associated with the synapse (Clarkson et al., 2010; Stankewich
et al., 2010). Thus, given that the Spna22/2 mice are deficient in
and mis-localize their ankyrins and bII- and bIII-spectrin, it is
likely that in these animals there are also problems with
membrane protein organization and sorting. In future studies it
will be important to examine this question in detail, both in vivo
and in cultured cells, to identify the full repertoire of proteins that
depend on spectrin and ankyrin for their efficient sorting to the
correct membrane compartment.
Materials and Methods
Generation of C57BL/6 (Spna22/2) mice
The embryonic stem (ES) cell line RRQ171 (BayGenomics, http://www.genetrap.
org) was generated by gene-trap protocol with pGTOLxf, containing a spliceacceptor sequence subcloned 59 end of a b-geo reporter cassette encoding a b-gal–
neomycin resistance fusion protein. ES cells heterozygous for the targeted
mutation were microinjected into C57BL/6 blastocysts and implanted into pseudopregnant foster mothers. Male chimeras were mated with C57BL/6 females. Ten
backcrosses established the gene-trapped allele on a C57BL/6 background. Mice
were genotyped by quantitative RT-PCR using primers bridging the b-geo
insertion and wild-type Spna2 sequence. The PCR primer sets used for detection
of the exon trap were: 59-CAAATGGCGATTACCGTTGA-39 and 39-GACAGTATCGGCCTCAGGAAGATCG-59. Spectrin primers used were exon 6 sense;
exon 7 reverse; exon 54 sense and exon 55 reverse. All animal experiments were
performed according to the relevant regulatory standards.
Analysis of mRNA and protein expression
Embryos were harvested from pregnant heterozygous mice typically at E13.5 to
E14.5. The levels of aII, bII, bIII, bIV, and bV–spectrin mRNA expression were
measured by real-time PCR, using primers summarized in supplementary material
Table S1. For western blot analysis, embryos were homogenized in 3.0 ml extract
buffer (20 mM HEPES pH 7.4, 120 mM NaCl, 25 mM KCl, 2 mM EDTA, 1 mM
EGTA, 1% Triton X-100) supplemented with Protease Arrest (Calbiochem)
(1:200). Supernatants were cleared at 28,000 g for 10 minutes at 4 ˚C, analyzed by
SDS-PAGE (NUPAGE Gel System, Invitrogen). Antibodies were mouse
monoclonal antibodies to aII-spectrin (Chemicon-clone 1622, and Santa Cruz
Biotechnology clone C-11), bII-spectrin (Pharmingen), b-actin clone AC-74
(Sigma), and b-galactosidase (Developmental Studies Hybridoma Bank). Rabbit
polyclonals were: aII-spectrin (Raf-A) and aI/bI-spectrin (Ras-C) (Harris et al.,
1985), ankyrin R (Cianci et al., 1988), bIII-spectrin (Stankewich et al., 1998), GFP
(Clontech), bIV-spectrin (a gift from Michele Solimena, Technical University,
Dresden, Germany). Other antibodies included anti-ankyrins B and G, rabbit antiVASP (sc-13975); mouse anti-pan14-3-3 (sc-1657); rabbit anti-Abi (sc-30038), all
from Santa Cruz Biotechnology. Transduction Labs provided mouse anti-KAP3
(K55520) and Cell Signaling provided mouse anti-Tau (4019). All antibodies were
used at 1:1000 dilutions for western blot analysis except that actin Mab was used at
1:20,000. For western blot analysis, protein loadings were calculated so as to
equalize the total protein load for each sample. This generally gave comparable
levels of actin loading, which served as an internal marker. Each lane of each
western blot represents a replicate sample from a different embryo.
Gradient sedimentation
Whole fresh E11.5 embryos were homogenized at 4 ˚C in 1 ml of homogenization
buffer (20 mM HEPES pH 7.4, 120 mM NaCl, 25 mM KCl, 2 mM EDTA, 1 mM
EGTA, 0.5% Triton X-100) supplemented with Protease Arrest (Calbiochem).
After centrifugation (1000 g for 10 minutes) to remove tissue debris, homogenates
were applied to a 5–20%, 13.5 ml continuous sucrose gradient and spun in a SW40 rotor (Beckman) at 40,000 rpm for 22.5 hours. Aliquots of 0.5 ml were
decanted from the top of the gradient mechanically using a gradient fractionator
(Labconco). The refractive index, which correlates with density, of all fractions
was analyzed with a refractometer.
Histology and immunolabeling
Immunoperoxidase staining of whole-mount embryos in paraffin sections was
carried out as before (Stankewich et al., 2006). Primary antibodies were used at a
1:500 dilution. For immunofluorescence, after the application of the primary
antibody, sections were incubated with either goat anti-mouse or anti-rabbit
secondary antibodies conjugated to Alexa Fluor dyes (Invitrogen) diluted 1:1000
for 1 hour at room temperature. To stain nuclei, slides were incubated in Hoechst
dye 33342 for 10 minutes. For visualization by alkaline phosphatase, the EnVision
G|2 reagent kit (Dako) was used in combination with Ferangi Blue (Biocare
Medical). Tissues were counterstained with Nuclear Fast Red (PolyScientific).
Slides were viewed by bright-field or fluorescent microscopy using an Olympus
AX70 microscope; image acquisition was processed using OpenLab software
(Improvision, Lexington, MA).
Analysis of primary mouse skin embryonic fibroblasts
Wild-type and Spna22/2 MEFs were harvested at E14.5 from embryos of
heterozygous crosses, and established in Dulbecco’s modified Eagle’s medium
(DMEM) supplemented with 10% fetal calf serum (FCS). All experiments were
duplicated in at least three independently established lines at passage numbers 3–9.
All cell lines were individually genotyped. For cell growth assays, 26104 MEF
cells were plated into each well of a 24-well Falcon tissue culture plate. Cells were
grown at 37 ˚C; at each time point cells were suspended in 0.05% trypsin–EDTA
and counted on an Improved Neubauer hemocytometer (each data point represents
an average of six samples). To test TGF-b1 stimulation, NIH3T3 cells and MEF
cells were grown and maintained in DMEM with 10% FBS (complete DMEM).
Cells were plated at 70–80% confluence in complete DMEM. After 24 hours,
culture media was exchanged with DMEM supplemented with only 1% FBS for
6 hours, then mouse TGF-b1 (Cell Signaling 5231) was added at 0, 2.5 or 10 ng/
ml for 24 and 48 hours. Cells were subsequently solubilized in gel loading buffer
and proteins analyzed by western blot analysis. Antibodies used were rabbit antiSmad2/3 and phospho-Smad2 (S465/467) (Cell Signaling s3102 and138D4,
respectively); mouse monoclonal anti-Smad3 (Santa Cruz Biotechnology sc101154); mouse monoclonal anti-Smad4 (Santa Cruz Biotechnology sc-7966); and
anti-Smad7, a gift from Mark Kidd (Yale University, New Haven, CT).
Cell spreading
MEFs were plated onto cover slips coated with 10 mg/ml fibronectin (Santa Cruz
sc-29011). After 1–3 hours, cells were fixed with 3.0% formaldehyde and stained
with Alexa Fluor 488–phallodin (Invitrogen). Images were captured and processed
as above. From each analysis, 500 cells were randomly selected for evaluation.
The relative area of the spread cells in both control and Spna22/2 MEFs was
measured using the Photoshop CS magic wand tool to score pixel number within
their well-defined cell borders. For recovery experiments, MEFs were transiently
transfected with Lipofectamine 2000 mixed with plasmid encoding fluorescently
tagged full-length aII-spectrin according to the manufacturer’s instructions
(Invitrogen). The pEYFP–aII-spectrin was prepared by subcloning a
Deletion of aII-spectrin
complementary DNA (cDNA)-encoding full-length human aII-spectrin (GenBank
U83867) into the pEYFP-C1 vector (Clontech), and was then shuttled into the
pPRIPu vector (Street et al., 2006).
Acknowledgements
We thank Michele Solimena for the bIV-spectrin antibody.
Funding
This study was supported in part by grants from the National
Institutes of Health [grant numbers R01-HL28560 and R01-DK
DK43812] to J.S.M. and by a grant from the National Ataxia
Foundation to M.C.S. Deposited in PMC for release after 12 months.
Supplementary material available online at
http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.080374/-/DC1
Journal of Cell Science
References
Bellenchi, G. C., Gurniak, C. B., Perlas, E., Middei, S., Ammassari-Teule, M. and
Witke, W. (2007). N-cofilin is associated with neuronal migration disorders and cell
cycle control in the cerebral cortex. Genes. Dev. 21, 2347-2357.
Bennett, V. and Baines, A. J. (2001). Spectrin and ankyrin-based pathways: metazoan
inventions for integrating cells into tissues. Physiol. Rev. 81, 1353-1392.
Bennett, V. and Healy, J. (2008). Organizing the fluid membrane bilayer: diseases
linked to spectrin and ankyrin. Trends Mol. Med. 14, 28-36.
Bialkowska, K., Saido, T. C. and Fox, J. E. (2005). SH3 domain of spectrin
participates in the activation of Rac in specialized calpain-induced integrin signaling
complexes. J. Cell Sci. 118, 381-395.
Bloch, R. J. and Morrow, J. S. (1989). An unusual b-spectrin associated with clustered
acetylcholine receptors. J. Cell Biol. 108, 481-493.
Bournier, O., Kroviarski, Y., Rotter, B., Nicolas, G., Lecomte, M. C. and Dhermy,
D. (2006). Spectrin interacts with EVL (Enabled/vasodilator-stimulated phosphoprotein-like protein), a protein involved in actin polymerization. Biol. Cell. 98, 279-293.
Chan, W., Kordeli, E. and Bennett, V. (1993). 440-kD ankyrinB: structure of the major
developmentally regulated domain and selective localization in unmyelinated axons.
J. Cell Biol. 123, 1463-1473.
Cianci, C. D., Giorgi, M. and Morrow, J. S. (1988). Phosphorylation of ankyrin downregulates its cooperative interaction with spectrin and protein 3. J. Cell Biochem. 37,
301-315.
Clarkson, Y. L., Gillespie, T., Perkins, E. M., Lyndon, A. R. and Jackson, M. (2010).
{beta}-III spectrin mutation L253P associated with spinocerebellar ataxia type 5
interferes with binding to Arp1 and protein trafficking from the Golgi. Hum. Mol.
Genet. 19, 3634-3641.
De Matteis, M. A. and Morrow, J. S. (2000). Spectrin tethers and mesh in the
biosynthetic pathway. J. Cell Sci. 113, 2331-2343.
Dubreuil, R. R. (2006). Functional links between membrane transport and the spectrin
cytoskeleton. J. Membr. Biol. 211, 151-161.
Eigenthaler, M., Engelhardt, S., Schinke, B., Kobsar, A., Schmitteckert, E.,
Gambaryan, S., Engelhardt, C. M., Krenn, V., Eliava, M., Jarchau, T. et al.
(2003). Disruption of cardiac Ena-VASP protein localization in intercalated disks
causes dilated cardiomyopathy. Am. J. Physiol. Heart Circ. Physiol. 285, H2471H2481.
Gallagher, P. G. and Jarolim, P. (2000). red cell membrane disorders. In Hematology:
Basic Principles and Practice, 5th edn (eds R. Hoffman, E. J. Benz, Jr, S. J. Shattil, B.
Furie, H. J. Cohen, L. E. Silberstein and P. McGlave), pp. 576-610. New York:
Churchill Livingstone.
Garbe, D. S., Das, A., Dubreuil, R. R. and Bashaw, G. J. (2007). beta-Spectrin
functions independently of Ankyrin to regulate the establishment and maintenance of
axon connections in the Drosophila embryonic CNS. Development 134, 273-284.
Golestaneh, N., Tang, Y., Katuri, V., Jogunoori, W., Mishra, L. and Mishra, B.
(2006). Cell cycle deregulation and loss of stem cell phenotype in the subventricular
zone of TGF-beta adaptor elf2/2 mouse brain. Brain Res. 1108, 45-53.
Gotz, M. and Huttner, W. B. (2005). The cell biology of neurogenesis. Nat. Rev. Mol.
Cell Biol. 6, 777-788.
Hammarlund, M., Jorgensen, E. M. and Bastiani, M. J. (2007). Axons break in
animals lacking beta-spectrin. J. Cell Biol. 176, 269-275.
Hanspal, M. and Palek, J. (1987). Synthesis and assembly of membrane skeletal
proteins in mammalian red cell precursors. J. Cell Biol. 105, 1417-1424.
Harris, A. S. and Morrow, J. S. (1990). Calmodulin and calcium-dependent protease I
coordinately regulate the interaction of fodrin with actin. Proc. Natl. Acad. Sci USA
87, 3009-3013.
Harris, A. S., Green, L. A., Ainger, K. J. and Morrow, J. S. (1985). Mechanism of
cytoskeletal regulation (I): functional differences correlate with antigenic dissimilarity
in human brain and erythrocyte spectrin. Biochim. Biophys. Acta 830, 147-158.
Holleran, E. A., Ligon, L. A., Tokito, M., Stankewich, M. C., Morrow, J. S. and
Holzbaur, E. L. (2001). beta III spectrin binds to the Arp1 subunit of dynactin. J.
Biol. Chem. 276, 36598-36605.
Hulsmeier, J., Pielage, J., Rickert, C., Technau, G. M., Klambt, C. and Stork, T.
(2007). Distinct functions of alpha-Spectrin and beta-Spectrin during axonal
pathfinding. Development 134, 713-722.
3965
Ikeda, Y., Dick, K. A., Weatherspoon, M. R., Gincel, D., Armbrust, K. R., Dalton,
J. C., Stevanin, G., Durr, A., Zuhlke, C., Burk, K. et al. (2006). Spectrin mutations
cause spinocerebellar ataxia type 5. Nat. Genet. 38, 184-190.
Kizhatil, K., Yoon, W., Mohler, P. J., Davis, L. H., Hoffman, J. A. and Bennett, V.
(2007). Ankyrin-G and beta2-spectrin collaborate in biogenesis of lateral membrane
of human bronchial epithelial cells. J. Biol. Chem. 282, 2029-2037.
Kordeli, E., Lambert, S. and Bennett, V. (1995). AnkyrinG. A new ankyrin gene with
neural-specific isoforms localized at the axonal initial segment and node of Ranvier.
J. Biol. Chem. 270, 2352-2359.
Lecomte, M. C., Nicolas, G., Fournier, C., Galand, C., Malbert-Colas, L., Bournier,
O., Dhermy, D. and Grandchamps, B. (2001). Interaction of spectrin with the low
molecular weight phosphotyrosine phosphatese (LMW-PTP). Cell Mol. Biol. Lett. 6,
216.
Li, D., Tang, H. Y. and Speicher, D. W. (2008). A structural model of the erythrocyte
spectrin heterodimer initiation site determined using homology modeling and
chemical cross-linking. J. Biol. Chem. 283, 1553-1562.
Luo, L. and O’Leary, D. D. (2005). Axon retraction and degeneration in development
and disease. Annu. Rev. Neurosci. 28, 127-156.
Mangeat, P. H. and Burridge, K. (1984). Immunoprecipitation of nonerythrocyte
spectrin within live cells following microinjection of specific antibodies: relation to
cytoskeletal structures. J. Cell Biol. 98, 1363-1377.
McMahon, L. W., Walsh, C. E. and Lambert, M. W. (1999). Human alpha spectrin II
and the fanconi anemia proteins FANCA and FANCC interact to form a nuclear
complex. J. Biol. Chem. 274, 32904-32908.
Metral, S., Machnicka, B., Bigot, S., Colin, Y., Dhermy, D. and Lecomte, M. C.
(2009). AlphaII-spectrin is critical for cell adhesion and cell cycle. J. Biol. Chem. 284,
2409-2418.
Mishra, L., Cai, T., Yu, P., Monga, S. P. and Mishra, B. (1999). Elf3 encodes a novel
200-kD beta-spectrin: role in liver development. Oncogene 18, 353-364.
Moorthy, S., Chen, L. and Bennett, V. (2000). Caenorhabditis elegans beta-G spectrin
is dispensable for establishment of epithelial polarity, but essential for muscular and
neuronal function. J. Cell Biol. 149, 915-930.
Muresan, V., Stankewich, M. C., Steffen, W., Morrow, J. S., Holzbaur, E. L. and
Schnapp, B. J. (2001). Dynactin-dependent, dynein-driven vesicle transport in the
absence of membrane proteins: a role for spectrin and acidic phospholipids. Mol. Cell
7, 173-183.
Nedrelow, J. H., Cianci, C. D. and Morrow, J. S. (2003). c-Src binds alpha II
spectrin’s Src homology 3 (SH3) domain and blocks calpain susceptibility by
phosphorylating Tyr1176. J. Biol. Chem. 278, 7735-7741.
Nicolas, G., Fournier, C. M., Galand, C., Malbert-Colas, L., Bournier, O.,
Kroviarski, Y., Bourgeois, M., Camonis, J. H., Dhermy, D., Grandchamp, B.
et al. (2002). Tyrosine phosphorylation regulates alpha II spectrin cleavage by
calpain. Mol. Cell Biol. 22, 3527-3536.
Perkins, E. M., Clarkson, Y. L., Sabatier, N., Longhurst, D. M., Millward, C. P.,
Jack, J., Toraiwa, J., Watanabe, M., Rothstein, J. D., Lyndon, A. R. et al. (2010).
Loss of beta-III spectrin leads to Purkinje cell dysfunction recapitulating the behavior
and neuropathology of spinocerebellar ataxia type 5 in humans. J. Neurosci. 30, 48574867.
Pradhan, D. and Morrow, J. (2002). The spectrin-ankyrin skeleton controls CD45
surface display and interleukin-2 production. Immunity 17, 303-315.
Pumplin, D. W. (1995). The membrane skeleton of acetylcholine receptor domains in
rat myotubes contains antiparallel homodimers of b-spectrin in filaments
quantitatively resembling those of erythrocytes. J. Cell Sci. 108, 3145-3154.
Qi, X., Yang, G., Yang, L., Lan, Y., Weng, T., Wang, J., Wu, Z., Xu, J., Gao, X. and
Yang, X. (2007). Essential role of Smad4 in maintaining cardiomyocyte proliferation
during murine embryonic heart development. Dev. Biol. 311, 136-146.
Ramser, E. M., Buck, F., Schachner, M. and Tilling, T. (2010). Binding of alphaII
spectrin to 14-3-3beta is involved in NCAM-dependent neurite outgrowth. Mol. Cell
Neurosci. 45, 66-74.
Sato, M. and Nagano, T. (2005). Involvement of filamin A and filamin A-interacting
protein (FILIP) in controlling the start and cell shape of radially migrating cortical
neurons. Anat. Sci. Int. 80, 19-29.
Stabach, P. R., Devarajan, P., Stankewich, M. C., Bannykh, S. and Morrow, J. S.
(2008). Ankyrin facilitates intracellular trafficking of alpha1-Na+-K+-ATPase in
polarized cells. Am. J. Physiol. Cell Physiol. 295, C1202-C1214.
Stabach, P. R., Simonovic, I., Ranieri, M. A., Aboodi, M. S., Steitz, T. A., Simonovic,
M. and Morrow, J. S. (2009). The structure of the ankyrin-binding site of betaspectrin reveals how tandem spectrin-repeats generate unique ligand-binding
properties. Blood 113, 5377-5384.
Stankewich, M. C., Stabach, P. R. and Morrow, J. S. (2006). Human Sec31B: a
family of new mammalian orthologues of yeast Sec31p that associate with the COPII
coat. J. Cell Sci. 119, 958-969.
Stankewich, M. C., Gwynn, B., Ardito, T., Ji, L., Kim, J., Robledo, R. F., Lux, S. E.,
Peters, L. L. and Morrow, J. S. (2010). Targeted deletion of betaIII spectrin impairs
synaptogenesis and generates ataxic and seizure phenotypes. Proc. Natl. Acad. Sci.
USA 107, 6022-6027.
Stankewich, M. C., Tse, W. T., Peters, L. L., Ch’ng, Y., John, K. M., Stabach, P. R.,
Devarajan, P., Morrow, J. S. and Lux, S. E. (1998). A widely expressed betaIII
spectrin associated with Golgi and cytoplasmic vesicles. Proc. Natl. Acad. Sci. USA
95, 14158-14163.
Street, M., Marsh, S. J., Stabach, P. R., Morrow, J. S., Brown, D. A. and Buckley,
N. J. (2006). Stimulation of Galphaq-coupled M1 muscarinic receptor causes
3966
Journal of Cell Science 124 (23)
Journal of Cell Science
reversible spectrin redistribution mediated by PLC, PKC and ROCK. J. Cell Sci. 119,
1528-1536.
Tahirovic, S., Hellal, F., Neukirchen, D., Hindges, R., Garvalov, B. K., Flynn, K. C.,
Stradal, T. E., Chrostek-Grashoff, A., Brakebusch, C. and Bradke, F. (2010).
Rac1 regulates neuronal polarization through the WAVE complex. J. Neurosci. 30,
6930-6943.
Takeda, S., Yamazaki, H., Seog, D. H., Kanai, Y., Terada, S. and Hirokawa, N.
(2000). Kinesin superfamily protein 3 (KIF3) motor transports fodrin-associating
vesicles important for neurite building. J. Cell Biol. 148, 1255-1265.
Tang, Y., Katuri, V., Iqbal, S., Narayan, T., Wang, Z., Lu, R. S., Mishra, L. and
Mishra, B. (2002). ELF a beta-spectrin is a neuronal precursor cell marker in
developing mammalian brain; structure and organization of the elf/beta-G spectrin
gene. Oncogene 21, 5255-5267.
Tang, Y., Katuri, V., Dillner, A., Mishra, B., Deng, C. X. and Mishra, L. (2003).
Disruption of transforming growth factor-beta signaling in ELF beta-spectrindeficient mice. Science 299, 574-577.
Tang, Y., Katuri, V., Srinivasan, R., Fogt, F., Redman, R., Anand, G., Said, A.,
Fishbein, T., Zasloff, M., Reddy, E. P. et al. (2005). Transforming growth factorbeta suppresses nonmetastatic colon cancer through Smad4 and adaptor protein ELF
at an early stage of tumorigenesis. Cancer Res. 65, 4228-4237.
Thomas, P. S., Kim, J., Nunez, S., Glogauer, M. and Kaartinen, V. (2010). Neural
crest cell-specific deletion of Rac1 results in defective cell-matrix interactions and
severe craniofacial and cardiovascular malformations. Dev. Biol. 340, 613-625.
Tse, W. T., Tang, J., Jin, O., Korsgren, C., John, K. M., Kung, A. L., Gwynn, B.,
Peters, L. L. and Lux, S. E. (2001). A new spectrin, beta IV, has a major truncated
isoform that associates with promyelocytic leukemia protein nuclear bodies and the
nuclear matrix. J. Biol. Chem. 276, 23974-23985.
Tuvia, S., Buhusi, M., Davis, L., Reedy, M. and Bennett, V. (1999). Ankyrin-B is
required for intracellular sorting of structurally diverse Ca2+ homeostasis proteins. J.
Cell Biol. 147, 995-1008.
Van Essen, D. C. (1997). A tension-based theory of morphogenesis and compact wiring
in the central nervous system. Nature 385, 313-318.
Woods, C. M. and Lazarides, E. (1985). Degradation of unassembled alpha- and betaspectrin by distinct intracellular pathways: regulation of spectrin topogenesis by betaspectrin degradation. Cell 40, 959-969.
Xu, J., Ziemnicka, D., Merz, G. S. and Kotula, L. (2000). Human spectrin Src
homology 3 domain binding protein 1 regulates macropinocytosis in NIH 3T3 cells. J.
Cell Sci. 113 Pt 21, 3805-3814.
Zhang, Y., Resneck, W. G., Lee, P. C., Randall, W. R., Bloch, R. J. and Ursitti, J. A.
(2010). Characterization and expression of a heart-selective alternatively spliced
variant of alpha II-spectrin, cardi+, during development in the rat. J. Mol. Cell
Cardiol. 48, 1050-1059.
Ziemnicka-Kotula, D., Xu, J., Gu, H., Potempska, A., Kim, K. S., Jenkins, E. C.,
Trenkner, E. and Kotula, L. (1998). Identification of a candidate human spectrin Src
homology 3 domain-binding protein suggests a general mechanism of association of
tyrosine kinases with the spectrin-based membrane skeleton. J. Biol. Chem. 273,
13681-13692.